Merge tag 'v4.4.121' into toradex_vf_4.4-nextColibri-VF_LXDE-Image_2.8b2.97-20180331

This is the 4.4.121 stable release

11 files changed, 727 insertions, 92 deletions

diff --git a/arch/x86/mm/Makefile b/arch/x86/mm/Makefile
index 65c47fda26fc..61e6cead9c4a 100644
--- a/arch/x86/mm/Makefile
+++ b/arch/x86/mm/Makefile
@@ -1,5 +1,5 @@
 obj-y	:=  init.o init_$(BITS).o fault.o ioremap.o extable.o pageattr.o mmap.o \
-	    pat.o pgtable.o physaddr.o gup.o setup_nx.o
+	    pat.o pgtable.o physaddr.o gup.o setup_nx.o tlb.o
 
 # Make sure __phys_addr has no stackprotector
 nostackp := $(call cc-option, -fno-stack-protector)
@@ -9,7 +9,6 @@ CFLAGS_setup_nx.o		:= $(nostackp)
 CFLAGS_fault.o := -I$(src)/../include/asm/trace
 
 obj-$(CONFIG_X86_PAT)		+= pat_rbtree.o
-obj-$(CONFIG_SMP)		+= tlb.o
 
 obj-$(CONFIG_X86_32)		+= pgtable_32.o iomap_32.o
 
@@ -33,3 +32,4 @@ obj-$(CONFIG_ACPI_NUMA)		+= srat.o
 obj-$(CONFIG_NUMA_EMU)		+= numa_emulation.o
 
 obj-$(CONFIG_X86_INTEL_MPX)	+= mpx.o
+obj-$(CONFIG_PAGE_TABLE_ISOLATION)		+= kaiser.o
diff --git a/arch/x86/mm/init.c b/arch/x86/mm/init.c
index 3aebbd6c6f5f..151fd33e9043 100644
--- a/arch/x86/mm/init.c
+++ b/arch/x86/mm/init.c
@@ -165,7 +165,7 @@ static void __init probe_page_size_mask(void)
 		cr4_set_bits_and_update_boot(X86_CR4_PSE);
 
 	/* Enable PGE if available */
-	if (cpu_has_pge) {
+	if (cpu_has_pge && !kaiser_enabled) {
 		cr4_set_bits_and_update_boot(X86_CR4_PGE);
 		__supported_pte_mask |= _PAGE_GLOBAL;
 	} else
@@ -753,13 +753,11 @@ void __init zone_sizes_init(void)
 }
 
 DEFINE_PER_CPU_SHARED_ALIGNED(struct tlb_state, cpu_tlbstate) = {
-#ifdef CONFIG_SMP
 	.active_mm = &init_mm,
 	.state = 0,
-#endif
 	.cr4 = ~0UL,	/* fail hard if we screw up cr4 shadow initialization */
 };
-EXPORT_SYMBOL_GPL(cpu_tlbstate);
+EXPORT_PER_CPU_SYMBOL(cpu_tlbstate);
 
 void update_cache_mode_entry(unsigned entry, enum page_cache_mode cache)
 {
diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c
index ec081fe0ce2c..d76ec9348cff 100644
--- a/arch/x86/mm/init_64.c
+++ b/arch/x86/mm/init_64.c
@@ -395,6 +395,16 @@ void __init cleanup_highmap(void)
 			continue;
 		if (vaddr < (unsigned long) _text || vaddr > end)
 			set_pmd(pmd, __pmd(0));
+		else if (kaiser_enabled) {
+			/*
+			 * level2_kernel_pgt is initialized with _PAGE_GLOBAL:
+			 * clear that now.  This is not important, so long as
+			 * CR4.PGE remains clear, but it removes an anomaly.
+			 * Physical mapping setup below avoids _PAGE_GLOBAL
+			 * by use of massage_pgprot() inside pfn_pte() etc.
+			 */
+			set_pmd(pmd, pmd_clear_flags(*pmd, _PAGE_GLOBAL));
+		}
 	}
 }
 
diff --git a/arch/x86/mm/ioremap.c b/arch/x86/mm/ioremap.c
index b9c78f3bcd67..53ab3f367472 100644
--- a/arch/x86/mm/ioremap.c
+++ b/arch/x86/mm/ioremap.c
@@ -348,11 +348,11 @@ void iounmap(volatile void __iomem *addr)
 	    (void __force *)addr < phys_to_virt(ISA_END_ADDRESS))
 		return;
 
+	mmiotrace_iounmap(addr);
+
 	addr = (volatile void __iomem *)
 		(PAGE_MASK & (unsigned long __force)addr);
 
-	mmiotrace_iounmap(addr);
-
 	/* Use the vm area unlocked, assuming the caller
 	   ensures there isn't another iounmap for the same address
 	   in parallel. Reuse of the virtual address is prevented by
diff --git a/arch/x86/mm/kaiser.c b/arch/x86/mm/kaiser.c
new file mode 100644
index 000000000000..7a72e32e4806
--- /dev/null
+++ b/arch/x86/mm/kaiser.c
@@ -0,0 +1,484 @@
+#include <linux/bug.h>
+#include <linux/kernel.h>
+#include <linux/errno.h>
+#include <linux/string.h>
+#include <linux/types.h>
+#include <linux/bug.h>
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/spinlock.h>
+#include <linux/mm.h>
+#include <linux/uaccess.h>
+#include <linux/ftrace.h>
+
+#undef pr_fmt
+#define pr_fmt(fmt)     "Kernel/User page tables isolation: " fmt
+
+#include <asm/kaiser.h>
+#include <asm/tlbflush.h>	/* to verify its kaiser declarations */
+#include <asm/pgtable.h>
+#include <asm/pgalloc.h>
+#include <asm/desc.h>
+#include <asm/cmdline.h>
+#include <asm/vsyscall.h>
+
+int kaiser_enabled __read_mostly = 1;
+EXPORT_SYMBOL(kaiser_enabled);	/* for inlined TLB flush functions */
+
+__visible
+DEFINE_PER_CPU_USER_MAPPED(unsigned long, unsafe_stack_register_backup);
+
+/*
+ * These can have bit 63 set, so we can not just use a plain "or"
+ * instruction to get their value or'd into CR3.  It would take
+ * another register.  So, we use a memory reference to these instead.
+ *
+ * This is also handy because systems that do not support PCIDs
+ * just end up or'ing a 0 into their CR3, which does no harm.
+ */
+DEFINE_PER_CPU(unsigned long, x86_cr3_pcid_user);
+
+/*
+ * At runtime, the only things we map are some things for CPU
+ * hotplug, and stacks for new processes.  No two CPUs will ever
+ * be populating the same addresses, so we only need to ensure
+ * that we protect between two CPUs trying to allocate and
+ * populate the same page table page.
+ *
+ * Only take this lock when doing a set_p[4um]d(), but it is not
+ * needed for doing a set_pte().  We assume that only the *owner*
+ * of a given allocation will be doing this for _their_
+ * allocation.
+ *
+ * This ensures that once a system has been running for a while
+ * and there have been stacks all over and these page tables
+ * are fully populated, there will be no further acquisitions of
+ * this lock.
+ */
+static DEFINE_SPINLOCK(shadow_table_allocation_lock);
+
+/*
+ * Returns -1 on error.
+ */
+static inline unsigned long get_pa_from_mapping(unsigned long vaddr)
+{
+	pgd_t *pgd;
+	pud_t *pud;
+	pmd_t *pmd;
+	pte_t *pte;
+
+	pgd = pgd_offset_k(vaddr);
+	/*
+	 * We made all the kernel PGDs present in kaiser_init().
+	 * We expect them to stay that way.
+	 */
+	BUG_ON(pgd_none(*pgd));
+	/*
+	 * PGDs are either 512GB or 128TB on all x86_64
+	 * configurations.  We don't handle these.
+	 */
+	BUG_ON(pgd_large(*pgd));
+
+	pud = pud_offset(pgd, vaddr);
+	if (pud_none(*pud)) {
+		WARN_ON_ONCE(1);
+		return -1;
+	}
+
+	if (pud_large(*pud))
+		return (pud_pfn(*pud) << PAGE_SHIFT) | (vaddr & ~PUD_PAGE_MASK);
+
+	pmd = pmd_offset(pud, vaddr);
+	if (pmd_none(*pmd)) {
+		WARN_ON_ONCE(1);
+		return -1;
+	}
+
+	if (pmd_large(*pmd))
+		return (pmd_pfn(*pmd) << PAGE_SHIFT) | (vaddr & ~PMD_PAGE_MASK);
+
+	pte = pte_offset_kernel(pmd, vaddr);
+	if (pte_none(*pte)) {
+		WARN_ON_ONCE(1);
+		return -1;
+	}
+
+	return (pte_pfn(*pte) << PAGE_SHIFT) | (vaddr & ~PAGE_MASK);
+}
+
+/*
+ * This is a relatively normal page table walk, except that it
+ * also tries to allocate page tables pages along the way.
+ *
+ * Returns a pointer to a PTE on success, or NULL on failure.
+ */
+static pte_t *kaiser_pagetable_walk(unsigned long address, bool user)
+{
+	pmd_t *pmd;
+	pud_t *pud;
+	pgd_t *pgd = native_get_shadow_pgd(pgd_offset_k(address));
+	gfp_t gfp = (GFP_KERNEL | __GFP_NOTRACK | __GFP_ZERO);
+	unsigned long prot = _KERNPG_TABLE;
+
+	if (pgd_none(*pgd)) {
+		WARN_ONCE(1, "All shadow pgds should have been populated");
+		return NULL;
+	}
+	BUILD_BUG_ON(pgd_large(*pgd) != 0);
+
+	if (user) {
+		/*
+		 * The vsyscall page is the only page that will have
+		 *  _PAGE_USER set. Catch everything else.
+		 */
+		BUG_ON(address != VSYSCALL_ADDR);
+
+		set_pgd(pgd, __pgd(pgd_val(*pgd) | _PAGE_USER));
+		prot = _PAGE_TABLE;
+	}
+
+	pud = pud_offset(pgd, address);
+	/* The shadow page tables do not use large mappings: */
+	if (pud_large(*pud)) {
+		WARN_ON(1);
+		return NULL;
+	}
+	if (pud_none(*pud)) {
+		unsigned long new_pmd_page = __get_free_page(gfp);
+		if (!new_pmd_page)
+			return NULL;
+		spin_lock(&shadow_table_allocation_lock);
+		if (pud_none(*pud)) {
+			set_pud(pud, __pud(prot | __pa(new_pmd_page)));
+			__inc_zone_page_state(virt_to_page((void *)
+						new_pmd_page), NR_KAISERTABLE);
+		} else
+			free_page(new_pmd_page);
+		spin_unlock(&shadow_table_allocation_lock);
+	}
+
+	pmd = pmd_offset(pud, address);
+	/* The shadow page tables do not use large mappings: */
+	if (pmd_large(*pmd)) {
+		WARN_ON(1);
+		return NULL;
+	}
+	if (pmd_none(*pmd)) {
+		unsigned long new_pte_page = __get_free_page(gfp);
+		if (!new_pte_page)
+			return NULL;
+		spin_lock(&shadow_table_allocation_lock);
+		if (pmd_none(*pmd)) {
+			set_pmd(pmd, __pmd(prot | __pa(new_pte_page)));
+			__inc_zone_page_state(virt_to_page((void *)
+						new_pte_page), NR_KAISERTABLE);
+		} else
+			free_page(new_pte_page);
+		spin_unlock(&shadow_table_allocation_lock);
+	}
+
+	return pte_offset_kernel(pmd, address);
+}
+
+static int kaiser_add_user_map(const void *__start_addr, unsigned long size,
+			       unsigned long flags)
+{
+	int ret = 0;
+	pte_t *pte;
+	unsigned long start_addr = (unsigned long )__start_addr;
+	unsigned long address = start_addr & PAGE_MASK;
+	unsigned long end_addr = PAGE_ALIGN(start_addr + size);
+	unsigned long target_address;
+
+	/*
+	 * It is convenient for callers to pass in __PAGE_KERNEL etc,
+	 * and there is no actual harm from setting _PAGE_GLOBAL, so
+	 * long as CR4.PGE is not set.  But it is nonetheless troubling
+	 * to see Kaiser itself setting _PAGE_GLOBAL (now that "nokaiser"
+	 * requires that not to be #defined to 0): so mask it off here.
+	 */
+	flags &= ~_PAGE_GLOBAL;
+	if (!(__supported_pte_mask & _PAGE_NX))
+		flags &= ~_PAGE_NX;
+
+	for (; address < end_addr; address += PAGE_SIZE) {
+		target_address = get_pa_from_mapping(address);
+		if (target_address == -1) {
+			ret = -EIO;
+			break;
+		}
+		pte = kaiser_pagetable_walk(address, flags & _PAGE_USER);
+		if (!pte) {
+			ret = -ENOMEM;
+			break;
+		}
+		if (pte_none(*pte)) {
+			set_pte(pte, __pte(flags | target_address));
+		} else {
+			pte_t tmp;
+			set_pte(&tmp, __pte(flags | target_address));
+			WARN_ON_ONCE(!pte_same(*pte, tmp));
+		}
+	}
+	return ret;
+}
+
+static int kaiser_add_user_map_ptrs(const void *start, const void *end, unsigned long flags)
+{
+	unsigned long size = end - start;
+
+	return kaiser_add_user_map(start, size, flags);
+}
+
+/*
+ * Ensure that the top level of the (shadow) page tables are
+ * entirely populated.  This ensures that all processes that get
+ * forked have the same entries.  This way, we do not have to
+ * ever go set up new entries in older processes.
+ *
+ * Note: we never free these, so there are no updates to them
+ * after this.
+ */
+static void __init kaiser_init_all_pgds(void)
+{
+	pgd_t *pgd;
+	int i = 0;
+
+	pgd = native_get_shadow_pgd(pgd_offset_k((unsigned long )0));
+	for (i = PTRS_PER_PGD / 2; i < PTRS_PER_PGD; i++) {
+		pgd_t new_pgd;
+		pud_t *pud = pud_alloc_one(&init_mm,
+					   PAGE_OFFSET + i * PGDIR_SIZE);
+		if (!pud) {
+			WARN_ON(1);
+			break;
+		}
+		inc_zone_page_state(virt_to_page(pud), NR_KAISERTABLE);
+		new_pgd = __pgd(_KERNPG_TABLE |__pa(pud));
+		/*
+		 * Make sure not to stomp on some other pgd entry.
+		 */
+		if (!pgd_none(pgd[i])) {
+			WARN_ON(1);
+			continue;
+		}
+		set_pgd(pgd + i, new_pgd);
+	}
+}
+
+#define kaiser_add_user_map_early(start, size, flags) do {	\
+	int __ret = kaiser_add_user_map(start, size, flags);	\
+	WARN_ON(__ret);						\
+} while (0)
+
+#define kaiser_add_user_map_ptrs_early(start, end, flags) do {		\
+	int __ret = kaiser_add_user_map_ptrs(start, end, flags);	\
+	WARN_ON(__ret);							\
+} while (0)
+
+void __init kaiser_check_boottime_disable(void)
+{
+	bool enable = true;
+	char arg[5];
+	int ret;
+
+	if (boot_cpu_has(X86_FEATURE_XENPV))
+		goto silent_disable;
+
+	ret = cmdline_find_option(boot_command_line, "pti", arg, sizeof(arg));
+	if (ret > 0) {
+		if (!strncmp(arg, "on", 2))
+			goto enable;
+
+		if (!strncmp(arg, "off", 3))
+			goto disable;
+
+		if (!strncmp(arg, "auto", 4))
+			goto skip;
+	}
+
+	if (cmdline_find_option_bool(boot_command_line, "nopti"))
+		goto disable;
+
+skip:
+	if (boot_cpu_data.x86_vendor == X86_VENDOR_AMD)
+		goto disable;
+
+enable:
+	if (enable)
+		setup_force_cpu_cap(X86_FEATURE_KAISER);
+
+	return;
+
+disable:
+	pr_info("disabled\n");
+
+silent_disable:
+	kaiser_enabled = 0;
+	setup_clear_cpu_cap(X86_FEATURE_KAISER);
+}
+
+/*
+ * If anything in here fails, we will likely die on one of the
+ * first kernel->user transitions and init will die.  But, we
+ * will have most of the kernel up by then and should be able to
+ * get a clean warning out of it.  If we BUG_ON() here, we run
+ * the risk of being before we have good console output.
+ */
+void __init kaiser_init(void)
+{
+	int cpu;
+
+	if (!kaiser_enabled)
+		return;
+
+	kaiser_init_all_pgds();
+
+	/*
+	 * Note that this sets _PAGE_USER and it needs to happen when the
+	 * pagetable hierarchy gets created, i.e., early. Otherwise
+	 * kaiser_pagetable_walk() will encounter initialized PTEs in the
+	 * hierarchy and not set the proper permissions, leading to the
+	 * pagefaults with page-protection violations when trying to read the
+	 * vsyscall page. For example.
+	 */
+	if (vsyscall_enabled())
+		kaiser_add_user_map_early((void *)VSYSCALL_ADDR,
+					  PAGE_SIZE,
+					  vsyscall_pgprot);
+
+	for_each_possible_cpu(cpu) {
+		void *percpu_vaddr = __per_cpu_user_mapped_start +
+				     per_cpu_offset(cpu);
+		unsigned long percpu_sz = __per_cpu_user_mapped_end -
+					  __per_cpu_user_mapped_start;
+		kaiser_add_user_map_early(percpu_vaddr, percpu_sz,
+					  __PAGE_KERNEL);
+	}
+
+	/*
+	 * Map the entry/exit text section, which is needed at
+	 * switches from user to and from kernel.
+	 */
+	kaiser_add_user_map_ptrs_early(__entry_text_start, __entry_text_end,
+				       __PAGE_KERNEL_RX);
+
+#ifdef CONFIG_FUNCTION_GRAPH_TRACER
+	kaiser_add_user_map_ptrs_early(__irqentry_text_start,
+				       __irqentry_text_end,
+				       __PAGE_KERNEL_RX);
+#endif
+	kaiser_add_user_map_early((void *)idt_descr.address,
+				  sizeof(gate_desc) * NR_VECTORS,
+				  __PAGE_KERNEL_RO);
+#ifdef CONFIG_TRACING
+	kaiser_add_user_map_early(&trace_idt_descr,
+				  sizeof(trace_idt_descr),
+				  __PAGE_KERNEL);
+	kaiser_add_user_map_early(&trace_idt_table,
+				  sizeof(gate_desc) * NR_VECTORS,
+				  __PAGE_KERNEL);
+#endif
+	kaiser_add_user_map_early(&debug_idt_descr, sizeof(debug_idt_descr),
+				  __PAGE_KERNEL);
+	kaiser_add_user_map_early(&debug_idt_table,
+				  sizeof(gate_desc) * NR_VECTORS,
+				  __PAGE_KERNEL);
+
+	pr_info("enabled\n");
+}
+
+/* Add a mapping to the shadow mapping, and synchronize the mappings */
+int kaiser_add_mapping(unsigned long addr, unsigned long size, unsigned long flags)
+{
+	if (!kaiser_enabled)
+		return 0;
+	return kaiser_add_user_map((const void *)addr, size, flags);
+}
+
+void kaiser_remove_mapping(unsigned long start, unsigned long size)
+{
+	extern void unmap_pud_range_nofree(pgd_t *pgd,
+				unsigned long start, unsigned long end);
+	unsigned long end = start + size;
+	unsigned long addr, next;
+	pgd_t *pgd;
+
+	if (!kaiser_enabled)
+		return;
+	pgd = native_get_shadow_pgd(pgd_offset_k(start));
+	for (addr = start; addr < end; pgd++, addr = next) {
+		next = pgd_addr_end(addr, end);
+		unmap_pud_range_nofree(pgd, addr, next);
+	}
+}
+
+/*
+ * Page table pages are page-aligned.  The lower half of the top
+ * level is used for userspace and the top half for the kernel.
+ * This returns true for user pages that need to get copied into
+ * both the user and kernel copies of the page tables, and false
+ * for kernel pages that should only be in the kernel copy.
+ */
+static inline bool is_userspace_pgd(pgd_t *pgdp)
+{
+	return ((unsigned long)pgdp % PAGE_SIZE) < (PAGE_SIZE / 2);
+}
+
+pgd_t kaiser_set_shadow_pgd(pgd_t *pgdp, pgd_t pgd)
+{
+	if (!kaiser_enabled)
+		return pgd;
+	/*
+	 * Do we need to also populate the shadow pgd?  Check _PAGE_USER to
+	 * skip cases like kexec and EFI which make temporary low mappings.
+	 */
+	if (pgd.pgd & _PAGE_USER) {
+		if (is_userspace_pgd(pgdp)) {
+			native_get_shadow_pgd(pgdp)->pgd = pgd.pgd;
+			/*
+			 * Even if the entry is *mapping* userspace, ensure
+			 * that userspace can not use it.  This way, if we
+			 * get out to userspace running on the kernel CR3,
+			 * userspace will crash instead of running.
+			 */
+			if (__supported_pte_mask & _PAGE_NX)
+				pgd.pgd |= _PAGE_NX;
+		}
+	} else if (!pgd.pgd) {
+		/*
+		 * pgd_clear() cannot check _PAGE_USER, and is even used to
+		 * clear corrupted pgd entries: so just rely on cases like
+		 * kexec and EFI never to be using pgd_clear().
+		 */
+		if (!WARN_ON_ONCE((unsigned long)pgdp & PAGE_SIZE) &&
+		    is_userspace_pgd(pgdp))
+			native_get_shadow_pgd(pgdp)->pgd = pgd.pgd;
+	}
+	return pgd;
+}
+
+void kaiser_setup_pcid(void)
+{
+	unsigned long user_cr3 = KAISER_SHADOW_PGD_OFFSET;
+
+	if (this_cpu_has(X86_FEATURE_PCID))
+		user_cr3 |= X86_CR3_PCID_USER_NOFLUSH;
+	/*
+	 * These variables are used by the entry/exit
+	 * code to change PCID and pgd and TLB flushing.
+	 */
+	this_cpu_write(x86_cr3_pcid_user, user_cr3);
+}
+
+/*
+ * Make a note that this cpu will need to flush USER tlb on return to user.
+ * If cpu does not have PCID, then the NOFLUSH bit will never have been set.
+ */
+void kaiser_flush_tlb_on_return_to_user(void)
+{
+	if (this_cpu_has(X86_FEATURE_PCID))
+		this_cpu_write(x86_cr3_pcid_user,
+			X86_CR3_PCID_USER_FLUSH | KAISER_SHADOW_PGD_OFFSET);
+}
+EXPORT_SYMBOL(kaiser_flush_tlb_on_return_to_user);
diff --git a/arch/x86/mm/kasan_init_64.c b/arch/x86/mm/kasan_init_64.c
index 4e5ac46adc9d..fdfa25c83119 100644
--- a/arch/x86/mm/kasan_init_64.c
+++ b/arch/x86/mm/kasan_init_64.c
@@ -121,11 +121,22 @@ void __init kasan_init(void)
 	kasan_populate_zero_shadow(kasan_mem_to_shadow((void *)MODULES_END),
 			(void *)KASAN_SHADOW_END);
 
-	memset(kasan_zero_page, 0, PAGE_SIZE);
-
 	load_cr3(init_level4_pgt);
 	__flush_tlb_all();
-	init_task.kasan_depth = 0;
 
+	/*
+	 * kasan_zero_page has been used as early shadow memory, thus it may
+	 * contain some garbage. Now we can clear and write protect it, since
+	 * after the TLB flush no one should write to it.
+	 */
+	memset(kasan_zero_page, 0, PAGE_SIZE);
+	for (i = 0; i < PTRS_PER_PTE; i++) {
+		pte_t pte = __pte(__pa(kasan_zero_page) | __PAGE_KERNEL_RO);
+		set_pte(&kasan_zero_pte[i], pte);
+	}
+	/* Flush TLBs again to be sure that write protection applied. */
+	__flush_tlb_all();
+
+	init_task.kasan_depth = 0;
 	pr_info("KernelAddressSanitizer initialized\n");
 }
diff --git a/arch/x86/mm/kmmio.c b/arch/x86/mm/kmmio.c
index ddb2244b06a1..76604c8a2a48 100644
--- a/arch/x86/mm/kmmio.c
+++ b/arch/x86/mm/kmmio.c
@@ -434,17 +434,18 @@ int register_kmmio_probe(struct kmmio_probe *p)
 	unsigned long flags;
 	int ret = 0;
 	unsigned long size = 0;
+	unsigned long addr = p->addr & PAGE_MASK;
 	const unsigned long size_lim = p->len + (p->addr & ~PAGE_MASK);
 	unsigned int l;
 	pte_t *pte;
 
 	spin_lock_irqsave(&kmmio_lock, flags);
-	if (get_kmmio_probe(p->addr)) {
+	if (get_kmmio_probe(addr)) {
 		ret = -EEXIST;
 		goto out;
 	}
 
-	pte = lookup_address(p->addr, &l);
+	pte = lookup_address(addr, &l);
 	if (!pte) {
 		ret = -EINVAL;
 		goto out;
@@ -453,7 +454,7 @@ int register_kmmio_probe(struct kmmio_probe *p)
 	kmmio_count++;
 	list_add_rcu(&p->list, &kmmio_probes);
 	while (size < size_lim) {
-		if (add_kmmio_fault_page(p->addr + size))
+		if (add_kmmio_fault_page(addr + size))
 			pr_err("Unable to set page fault.\n");
 		size += page_level_size(l);
 	}
@@ -527,19 +528,20 @@ void unregister_kmmio_probe(struct kmmio_probe *p)
 {
 	unsigned long flags;
 	unsigned long size = 0;
+	unsigned long addr = p->addr & PAGE_MASK;
 	const unsigned long size_lim = p->len + (p->addr & ~PAGE_MASK);
 	struct kmmio_fault_page *release_list = NULL;
 	struct kmmio_delayed_release *drelease;
 	unsigned int l;
 	pte_t *pte;
 
-	pte = lookup_address(p->addr, &l);
+	pte = lookup_address(addr, &l);
 	if (!pte)
 		return;
 
 	spin_lock_irqsave(&kmmio_lock, flags);
 	while (size < size_lim) {
-		release_kmmio_fault_page(p->addr + size, &release_list);
+		release_kmmio_fault_page(addr + size, &release_list);
 		size += page_level_size(l);
 	}
 	list_del_rcu(&p->list);
diff --git a/arch/x86/mm/pageattr.c b/arch/x86/mm/pageattr.c
index b599a780a5a9..79377e2a7bcd 100644
--- a/arch/x86/mm/pageattr.c
+++ b/arch/x86/mm/pageattr.c
@@ -52,6 +52,7 @@ static DEFINE_SPINLOCK(cpa_lock);
 #define CPA_FLUSHTLB 1
 #define CPA_ARRAY 2
 #define CPA_PAGES_ARRAY 4
+#define CPA_FREE_PAGETABLES 8
 
 #ifdef CONFIG_PROC_FS
 static unsigned long direct_pages_count[PG_LEVEL_NUM];
@@ -723,10 +724,13 @@ static int split_large_page(struct cpa_data *cpa, pte_t *kpte,
 	return 0;
 }
 
-static bool try_to_free_pte_page(pte_t *pte)
+static bool try_to_free_pte_page(struct cpa_data *cpa, pte_t *pte)
 {
 	int i;
 
+	if (!(cpa->flags & CPA_FREE_PAGETABLES))
+		return false;
+
 	for (i = 0; i < PTRS_PER_PTE; i++)
 		if (!pte_none(pte[i]))
 			return false;
@@ -735,10 +739,13 @@ static bool try_to_free_pte_page(pte_t *pte)
 	return true;
 }
 
-static bool try_to_free_pmd_page(pmd_t *pmd)
+static bool try_to_free_pmd_page(struct cpa_data *cpa, pmd_t *pmd)
 {
 	int i;
 
+	if (!(cpa->flags & CPA_FREE_PAGETABLES))
+		return false;
+
 	for (i = 0; i < PTRS_PER_PMD; i++)
 		if (!pmd_none(pmd[i]))
 			return false;
@@ -759,7 +766,9 @@ static bool try_to_free_pud_page(pud_t *pud)
 	return true;
 }
 
-static bool unmap_pte_range(pmd_t *pmd, unsigned long start, unsigned long end)
+static bool unmap_pte_range(struct cpa_data *cpa, pmd_t *pmd,
+			    unsigned long start,
+			    unsigned long end)
 {
 	pte_t *pte = pte_offset_kernel(pmd, start);
 
@@ -770,22 +779,23 @@ static bool unmap_pte_range(pmd_t *pmd, unsigned long start, unsigned long end)
 		pte++;
 	}
 
-	if (try_to_free_pte_page((pte_t *)pmd_page_vaddr(*pmd))) {
+	if (try_to_free_pte_page(cpa, (pte_t *)pmd_page_vaddr(*pmd))) {
 		pmd_clear(pmd);
 		return true;
 	}
 	return false;
 }
 
-static void __unmap_pmd_range(pud_t *pud, pmd_t *pmd,
+static void __unmap_pmd_range(struct cpa_data *cpa, pud_t *pud, pmd_t *pmd,
 			      unsigned long start, unsigned long end)
 {
-	if (unmap_pte_range(pmd, start, end))
-		if (try_to_free_pmd_page((pmd_t *)pud_page_vaddr(*pud)))
+	if (unmap_pte_range(cpa, pmd, start, end))
+		if (try_to_free_pmd_page(cpa, (pmd_t *)pud_page_vaddr(*pud)))
 			pud_clear(pud);
 }
 
-static void unmap_pmd_range(pud_t *pud, unsigned long start, unsigned long end)
+static void unmap_pmd_range(struct cpa_data *cpa, pud_t *pud,
+			    unsigned long start, unsigned long end)
 {
 	pmd_t *pmd = pmd_offset(pud, start);
 
@@ -796,7 +806,7 @@ static void unmap_pmd_range(pud_t *pud, unsigned long start, unsigned long end)
 		unsigned long next_page = (start + PMD_SIZE) & PMD_MASK;
 		unsigned long pre_end = min_t(unsigned long, end, next_page);
 
-		__unmap_pmd_range(pud, pmd, start, pre_end);
+		__unmap_pmd_range(cpa, pud, pmd, start, pre_end);
 
 		start = pre_end;
 		pmd++;
@@ -809,7 +819,8 @@ static void unmap_pmd_range(pud_t *pud, unsigned long start, unsigned long end)
 		if (pmd_large(*pmd))
 			pmd_clear(pmd);
 		else
-			__unmap_pmd_range(pud, pmd, start, start + PMD_SIZE);
+			__unmap_pmd_range(cpa, pud, pmd,
+					  start, start + PMD_SIZE);
 
 		start += PMD_SIZE;
 		pmd++;
@@ -819,17 +830,19 @@ static void unmap_pmd_range(pud_t *pud, unsigned long start, unsigned long end)
 	 * 4K leftovers?
 	 */
 	if (start < end)
-		return __unmap_pmd_range(pud, pmd, start, end);
+		return __unmap_pmd_range(cpa, pud, pmd, start, end);
 
 	/*
 	 * Try again to free the PMD page if haven't succeeded above.
 	 */
 	if (!pud_none(*pud))
-		if (try_to_free_pmd_page((pmd_t *)pud_page_vaddr(*pud)))
+		if (try_to_free_pmd_page(cpa, (pmd_t *)pud_page_vaddr(*pud)))
 			pud_clear(pud);
 }
 
-static void unmap_pud_range(pgd_t *pgd, unsigned long start, unsigned long end)
+static void __unmap_pud_range(struct cpa_data *cpa, pgd_t *pgd,
+			      unsigned long start,
+			      unsigned long end)
 {
 	pud_t *pud = pud_offset(pgd, start);
 
@@ -840,7 +853,7 @@ static void unmap_pud_range(pgd_t *pgd, unsigned long start, unsigned long end)
 		unsigned long next_page = (start + PUD_SIZE) & PUD_MASK;
 		unsigned long pre_end	= min_t(unsigned long, end, next_page);
 
-		unmap_pmd_range(pud, start, pre_end);
+		unmap_pmd_range(cpa, pud, start, pre_end);
 
 		start = pre_end;
 		pud++;
@@ -854,7 +867,7 @@ static void unmap_pud_range(pgd_t *pgd, unsigned long start, unsigned long end)
 		if (pud_large(*pud))
 			pud_clear(pud);
 		else
-			unmap_pmd_range(pud, start, start + PUD_SIZE);
+			unmap_pmd_range(cpa, pud, start, start + PUD_SIZE);
 
 		start += PUD_SIZE;
 		pud++;
@@ -864,7 +877,7 @@ static void unmap_pud_range(pgd_t *pgd, unsigned long start, unsigned long end)
 	 * 2M leftovers?
 	 */
 	if (start < end)
-		unmap_pmd_range(pud, start, end);
+		unmap_pmd_range(cpa, pud, start, end);
 
 	/*
 	 * No need to try to free the PUD page because we'll free it in
@@ -872,6 +885,24 @@ static void unmap_pud_range(pgd_t *pgd, unsigned long start, unsigned long end)
 	 */
 }
 
+static void unmap_pud_range(pgd_t *pgd, unsigned long start, unsigned long end)
+{
+	struct cpa_data cpa = {
+		.flags = CPA_FREE_PAGETABLES,
+	};
+
+	__unmap_pud_range(&cpa, pgd, start, end);
+}
+
+void unmap_pud_range_nofree(pgd_t *pgd, unsigned long start, unsigned long end)
+{
+	struct cpa_data cpa = {
+		.flags = 0,
+	};
+
+	__unmap_pud_range(&cpa, pgd, start, end);
+}
+
 static void unmap_pgd_range(pgd_t *root, unsigned long addr, unsigned long end)
 {
 	pgd_t *pgd_entry = root + pgd_index(addr);
diff --git a/arch/x86/mm/pat.c b/arch/x86/mm/pat.c
index 3f1bb4f93a5a..3146b1da6d72 100644
--- a/arch/x86/mm/pat.c
+++ b/arch/x86/mm/pat.c
@@ -750,11 +750,8 @@ static inline int range_is_allowed(unsigned long pfn, unsigned long size)
 		return 1;
 
 	while (cursor < to) {
-		if (!devmem_is_allowed(pfn)) {
-			pr_info("x86/PAT: Program %s tried to access /dev/mem between [mem %#010Lx-%#010Lx], PAT prevents it\n",
-				current->comm, from, to - 1);
+		if (!devmem_is_allowed(pfn))
 			return 0;
-		}
 		cursor += PAGE_SIZE;
 		pfn++;
 	}
diff --git a/arch/x86/mm/pgtable.c b/arch/x86/mm/pgtable.c
index fb0a9dd1d6e4..dbc27a2b4ad5 100644
--- a/arch/x86/mm/pgtable.c
+++ b/arch/x86/mm/pgtable.c
@@ -6,7 +6,7 @@
 #include <asm/fixmap.h>
 #include <asm/mtrr.h>
 
-#define PGALLOC_GFP GFP_KERNEL | __GFP_NOTRACK | __GFP_REPEAT | __GFP_ZERO
+#define PGALLOC_GFP (GFP_KERNEL | __GFP_NOTRACK | __GFP_REPEAT | __GFP_ZERO)
 
 #ifdef CONFIG_HIGHPTE
 #define PGALLOC_USER_GFP __GFP_HIGHMEM
@@ -340,14 +340,24 @@ static inline void _pgd_free(pgd_t *pgd)
 		kmem_cache_free(pgd_cache, pgd);
 }
 #else
+
+/*
+ * Instead of one pgd, Kaiser acquires two pgds.  Being order-1, it is
+ * both 8k in size and 8k-aligned.  That lets us just flip bit 12
+ * in a pointer to swap between the two 4k halves.
+ */
+#define PGD_ALLOCATION_ORDER	kaiser_enabled
+
 static inline pgd_t *_pgd_alloc(void)
 {
-	return (pgd_t *)__get_free_page(PGALLOC_GFP);
+	/* No __GFP_REPEAT: to avoid page allocation stalls in order-1 case */
+	return (pgd_t *)__get_free_pages(PGALLOC_GFP & ~__GFP_REPEAT,
+					 PGD_ALLOCATION_ORDER);
 }
 
 static inline void _pgd_free(pgd_t *pgd)
 {
-	free_page((unsigned long)pgd);
+	free_pages((unsigned long)pgd, PGD_ALLOCATION_ORDER);
 }
 #endif /* CONFIG_X86_PAE */
 
diff --git a/arch/x86/mm/tlb.c b/arch/x86/mm/tlb.c
index 5a760fd66bec..7cad01af6dcd 100644
--- a/arch/x86/mm/tlb.c
+++ b/arch/x86/mm/tlb.c
@@ -6,16 +6,17 @@
 #include <linux/interrupt.h>
 #include <linux/module.h>
 #include <linux/cpu.h>
+#include <linux/debugfs.h>
 
 #include <asm/tlbflush.h>
 #include <asm/mmu_context.h>
 #include <asm/cache.h>
 #include <asm/apic.h>
 #include <asm/uv/uv.h>
-#include <linux/debugfs.h>
+#include <asm/kaiser.h>
 
 /*
- *	Smarter SMP flushing macros.
+ *	TLB flushing, formerly SMP-only
  *		c/o Linus Torvalds.
  *
  *	These mean you can really definitely utterly forget about
@@ -34,6 +35,36 @@ struct flush_tlb_info {
 	unsigned long flush_end;
 };
 
+static void load_new_mm_cr3(pgd_t *pgdir)
+{
+	unsigned long new_mm_cr3 = __pa(pgdir);
+
+	if (kaiser_enabled) {
+		/*
+		 * We reuse the same PCID for different tasks, so we must
+		 * flush all the entries for the PCID out when we change tasks.
+		 * Flush KERN below, flush USER when returning to userspace in
+		 * kaiser's SWITCH_USER_CR3 (_SWITCH_TO_USER_CR3) macro.
+		 *
+		 * invpcid_flush_single_context(X86_CR3_PCID_ASID_USER) could
+		 * do it here, but can only be used if X86_FEATURE_INVPCID is
+		 * available - and many machines support pcid without invpcid.
+		 *
+		 * If X86_CR3_PCID_KERN_FLUSH actually added something, then it
+		 * would be needed in the write_cr3() below - if PCIDs enabled.
+		 */
+		BUILD_BUG_ON(X86_CR3_PCID_KERN_FLUSH);
+		kaiser_flush_tlb_on_return_to_user();
+	}
+
+	/*
+	 * Caution: many callers of this function expect
+	 * that load_cr3() is serializing and orders TLB
+	 * fills with respect to the mm_cpumask writes.
+	 */
+	write_cr3(new_mm_cr3);
+}
+
 /*
  * We cannot call mmdrop() because we are in interrupt context,
  * instead update mm->cpu_vm_mask.
@@ -45,7 +76,7 @@ void leave_mm(int cpu)
 		BUG();
 	if (cpumask_test_cpu(cpu, mm_cpumask(active_mm))) {
 		cpumask_clear_cpu(cpu, mm_cpumask(active_mm));
-		load_cr3(swapper_pg_dir);
+		load_new_mm_cr3(swapper_pg_dir);
 		/*
 		 * This gets called in the idle path where RCU
 		 * functions differently.  Tracing normally
@@ -57,6 +88,109 @@ void leave_mm(int cpu)
 }
 EXPORT_SYMBOL_GPL(leave_mm);
 
+void switch_mm(struct mm_struct *prev, struct mm_struct *next,
+	       struct task_struct *tsk)
+{
+	unsigned long flags;
+
+	local_irq_save(flags);
+	switch_mm_irqs_off(prev, next, tsk);
+	local_irq_restore(flags);
+}
+
+void switch_mm_irqs_off(struct mm_struct *prev, struct mm_struct *next,
+			struct task_struct *tsk)
+{
+	unsigned cpu = smp_processor_id();
+
+	if (likely(prev != next)) {
+		this_cpu_write(cpu_tlbstate.state, TLBSTATE_OK);
+		this_cpu_write(cpu_tlbstate.active_mm, next);
+		cpumask_set_cpu(cpu, mm_cpumask(next));
+
+		/*
+		 * Re-load page tables.
+		 *
+		 * This logic has an ordering constraint:
+		 *
+		 *  CPU 0: Write to a PTE for 'next'
+		 *  CPU 0: load bit 1 in mm_cpumask.  if nonzero, send IPI.
+		 *  CPU 1: set bit 1 in next's mm_cpumask
+		 *  CPU 1: load from the PTE that CPU 0 writes (implicit)
+		 *
+		 * We need to prevent an outcome in which CPU 1 observes
+		 * the new PTE value and CPU 0 observes bit 1 clear in
+		 * mm_cpumask.  (If that occurs, then the IPI will never
+		 * be sent, and CPU 0's TLB will contain a stale entry.)
+		 *
+		 * The bad outcome can occur if either CPU's load is
+		 * reordered before that CPU's store, so both CPUs must
+		 * execute full barriers to prevent this from happening.
+		 *
+		 * Thus, switch_mm needs a full barrier between the
+		 * store to mm_cpumask and any operation that could load
+		 * from next->pgd.  TLB fills are special and can happen
+		 * due to instruction fetches or for no reason at all,
+		 * and neither LOCK nor MFENCE orders them.
+		 * Fortunately, load_cr3() is serializing and gives the
+		 * ordering guarantee we need.
+		 *
+		 */
+		load_new_mm_cr3(next->pgd);
+
+		trace_tlb_flush(TLB_FLUSH_ON_TASK_SWITCH, TLB_FLUSH_ALL);
+
+		/* Stop flush ipis for the previous mm */
+		cpumask_clear_cpu(cpu, mm_cpumask(prev));
+
+		/* Load per-mm CR4 state */
+		load_mm_cr4(next);
+
+#ifdef CONFIG_MODIFY_LDT_SYSCALL
+		/*
+		 * Load the LDT, if the LDT is different.
+		 *
+		 * It's possible that prev->context.ldt doesn't match
+		 * the LDT register.  This can happen if leave_mm(prev)
+		 * was called and then modify_ldt changed
+		 * prev->context.ldt but suppressed an IPI to this CPU.
+		 * In this case, prev->context.ldt != NULL, because we
+		 * never set context.ldt to NULL while the mm still
+		 * exists.  That means that next->context.ldt !=
+		 * prev->context.ldt, because mms never share an LDT.
+		 */
+		if (unlikely(prev->context.ldt != next->context.ldt))
+			load_mm_ldt(next);
+#endif
+	} else {
+		this_cpu_write(cpu_tlbstate.state, TLBSTATE_OK);
+		BUG_ON(this_cpu_read(cpu_tlbstate.active_mm) != next);
+
+		if (!cpumask_test_cpu(cpu, mm_cpumask(next))) {
+			/*
+			 * On established mms, the mm_cpumask is only changed
+			 * from irq context, from ptep_clear_flush() while in
+			 * lazy tlb mode, and here. Irqs are blocked during
+			 * schedule, protecting us from simultaneous changes.
+			 */
+			cpumask_set_cpu(cpu, mm_cpumask(next));
+
+			/*
+			 * We were in lazy tlb mode and leave_mm disabled
+			 * tlb flush IPI delivery. We must reload CR3
+			 * to make sure to use no freed page tables.
+			 *
+			 * As above, load_cr3() is serializing and orders TLB
+			 * fills with respect to the mm_cpumask write.
+			 */
+			load_new_mm_cr3(next->pgd);
+			trace_tlb_flush(TLB_FLUSH_ON_TASK_SWITCH, TLB_FLUSH_ALL);
+			load_mm_cr4(next);
+			load_mm_ldt(next);
+		}
+	}
+}
+
 /*
  * The flush IPI assumes that a thread switch happens in this order:
  * [cpu0: the cpu that switches]
@@ -104,7 +238,7 @@ static void flush_tlb_func(void *info)
 
 	inc_irq_stat(irq_tlb_count);
 
-	if (f->flush_mm != this_cpu_read(cpu_tlbstate.active_mm))
+	if (f->flush_mm && f->flush_mm != this_cpu_read(cpu_tlbstate.active_mm))
 		return;
 
 	count_vm_tlb_event(NR_TLB_REMOTE_FLUSH_RECEIVED);
@@ -158,23 +292,6 @@ void native_flush_tlb_others(const struct cpumask *cpumask,
 	smp_call_function_many(cpumask, flush_tlb_func, &info, 1);
 }
 
-void flush_tlb_current_task(void)
-{
-	struct mm_struct *mm = current->mm;
-
-	preempt_disable();
-
-	count_vm_tlb_event(NR_TLB_LOCAL_FLUSH_ALL);
-
-	/* This is an implicit full barrier that synchronizes with switch_mm. */
-	local_flush_tlb();
-
-	trace_tlb_flush(TLB_LOCAL_SHOOTDOWN, TLB_FLUSH_ALL);
-	if (cpumask_any_but(mm_cpumask(mm), smp_processor_id()) < nr_cpu_ids)
-		flush_tlb_others(mm_cpumask(mm), mm, 0UL, TLB_FLUSH_ALL);
-	preempt_enable();
-}
-
 /*
  * See Documentation/x86/tlb.txt for details.  We choose 33
  * because it is large enough to cover the vast majority (at
@@ -195,6 +312,12 @@ void flush_tlb_mm_range(struct mm_struct *mm, unsigned long start,
 	unsigned long base_pages_to_flush = TLB_FLUSH_ALL;
 
 	preempt_disable();
+
+	if ((end != TLB_FLUSH_ALL) && !(vmflag & VM_HUGETLB))
+		base_pages_to_flush = (end - start) >> PAGE_SHIFT;
+	if (base_pages_to_flush > tlb_single_page_flush_ceiling)
+		base_pages_to_flush = TLB_FLUSH_ALL;
+
 	if (current->active_mm != mm) {
 		/* Synchronize with switch_mm. */
 		smp_mb();
@@ -211,15 +334,11 @@ void flush_tlb_mm_range(struct mm_struct *mm, unsigned long start,
 		goto out;
 	}
 
-	if ((end != TLB_FLUSH_ALL) && !(vmflag & VM_HUGETLB))
-		base_pages_to_flush = (end - start) >> PAGE_SHIFT;
-
 	/*
 	 * Both branches below are implicit full barriers (MOV to CR or
 	 * INVLPG) that synchronize with switch_mm.
 	 */
-	if (base_pages_to_flush > tlb_single_page_flush_ceiling) {
-		base_pages_to_flush = TLB_FLUSH_ALL;
+	if (base_pages_to_flush == TLB_FLUSH_ALL) {
 		count_vm_tlb_event(NR_TLB_LOCAL_FLUSH_ALL);
 		local_flush_tlb();
 	} else {
@@ -240,33 +359,6 @@ out:
 	preempt_enable();
 }
 
-void flush_tlb_page(struct vm_area_struct *vma, unsigned long start)
-{
-	struct mm_struct *mm = vma->vm_mm;
-
-	preempt_disable();
-
-	if (current->active_mm == mm) {
-		if (current->mm) {
-			/*
-			 * Implicit full barrier (INVLPG) that synchronizes
-			 * with switch_mm.
-			 */
-			__flush_tlb_one(start);
-		} else {
-			leave_mm(smp_processor_id());
-
-			/* Synchronize with switch_mm. */
-			smp_mb();
-		}
-	}
-
-	if (cpumask_any_but(mm_cpumask(mm), smp_processor_id()) < nr_cpu_ids)
-		flush_tlb_others(mm_cpumask(mm), mm, start, start + PAGE_SIZE);
-
-	preempt_enable();
-}
-
 static void do_flush_tlb_all(void *info)
 {
 	count_vm_tlb_event(NR_TLB_REMOTE_FLUSH_RECEIVED);